SASITe32


 * SASIT E32: A European Union Sentence Repetition Task to Measure Acquisition of Increasingly Complex Language Structures **

Mostly I put this up here because I kept forgetting the name of the test ... and it would sometimes take me a while to find it. I'm planning to mess around with this instrument this year to see if it contributes towards measuring English Learners' acquisition of language structures.

So you’re interested in an instrument to make sure we’re not “false identifying” kids with lower literacy levels (because of instructional or poverty-related reasons) as ESL, just because they’re getting lower scores on the Reading and Writing sections of our screeners and ACCESS test. Basically, you’re wondering if kids are getting “trapped” in ESL because of non-SLA poor literacy skills.


 * ** The website for the entire SASIT Project ** || [] ||
 * Erlam's Ch 3. on Sentence Repetition tasks to measure "implicit learning" || This is Rosemary Erlam's article, "The Elicited Oral Imitation Test as a Measure of Implicit Knowledge."

Basically, it's the clearest, most well written source I know on using sentence repetition as a test of the acquisition of grammatical structures in SLA. It's also Chapter 3 from Rod Ellis's 2009 //Implicit and Explicit Knowledge in Second Language Learning, Testing and Teaching//

//I'm going to link this because this is simply the best introduction to this entire issue I know of .. I'll track Dr. Erlam down and ask her permission ...// || to have to convert it to American English before I can use it.
 * Marinis et al on Sentence Repetition tasks || This is the theoretical article that introduces the task ||
 * [[file:SASIT-E32-Protocol,-Structures-and-Scoresheet (1).pdf|SASIT Protocol]] || This is the SASIT protocol itself from a UK website; I'm going

//I'm not certain I'm supposed to have this floating around the// //Internet. They did ... but if they don't want it there, they can email// me ... john.wolfe.mps at Gmail. ||
 * **SASITe32 re-done in American English**

| || I yankee-doodled this thing -- but I didn't check to make sure the points work out exactly, but I figure it's close enough. || * ( The goal is to experience whether a mere repetition task could serve as a decent indicator of whether or not students have learned a grammatical structure, etc. Can you repeat phonetically -- or do you have to have a handle on the language in order to repeat?)
 * A totally amateur version of an Elicited Imitation /Sentence Repetition task in Spanish. || Experience the amateur Spanish sentence repetition task I played around with: [|Here].


 * [[image:little-test.jpg link="@http://tiny.cc/jjwSpanishRepeat"]] ||

This might be the type of sentence repetition task that could check for false positives**. Basically, this test should measure “acquired syntax.” The idea is that to be really proficient in English, you have to acquire a lot of syntactic structures. Native speakers get them as a free gift, but acquiring them is more effortful for English Learners. (This test also has some sentences that prove difficult for all kids – because of pronoun ambiguity, less common structures, or sentences that stress out working memory or the acoustic loop or whatever. But still….)**


 * Here’s the recently developed “ **School-Age Sentence Imitation Test-E32” – or the SASIT -E32. **(This is similar to the OLAI test, but I’m pretty sure this one is still in an early, free, development stage. It hasn’t been totally taken over by the publishers the way the OLAI has..)**


 * It seems to be designed as part of a large-scale European effort to develop diagnostics for SLI’s (“Specific Language Impairments”) in multilingual settings. (The test is written in British English, so some of the sentences would have to be revised) **


 * The point is that it does offer a tested and validated set of sentences of “increasing morphosyntactic complexity.” **


 * This actually starts at a fairly complex sentence level – with an early sentence like: **


 * “She was seen by the doctor in the morning.” **


 * But the real complexity comes towards the end, with lots of verb shifts, embedding and recursion: **


 * “The horse that the farmer pushed kicked him in the back.” **


 * At the same time, the vocabulary never gets that tough. Unlike with the OLAI, the sentences in this test are produced and repeated without picture support. And they’re fairly de-contextualized. They don’t add up to a story. I think the reason is that this ends up working as a type of “stress test” – structures that support linguistic and mnemonic functions are removed, so that success in the repetition task really relies on having the language structures down pat. **


 * In the second PDF – the article about the test – you’ll see examples of visuals that were developed to keep track of and motivate the test takers. (You see a bear moving down a path for every sentence the student repeats, for example.) **

IF YOU WANTED TO USE THIS TO INVESTIGATE WHETHER THE ACCESS IS PRODUCING FALSE POSITIVES,** here’s one way to do it:

 o If the native English kid performs considerably stronger on the SASET-E32, then the implication is that the seemingly similar reading levels are “masking” quite different language acquisition levels.  o If, on the other hand, the English Learner performs about the same as the native English kid, it could suggest that a relatively low literacy level is “trapping” a kid in ESL status despite higher acquisition levels (as indicated by what’s basically a syntax acquisition task).
 * Find kids (native English and let’s say Level 4 WIDA) who are scoring roughly the same on some literacy measure (like F&P)
 * Administer the test.

**What I would expect to see** is acquisition-driven differences in ability showing up early and dramatically. I’m thinking that native English speakers would do A LOT better at this – but I could be wrong.

**If the differences show up early enough,** you could get by with a shorter version of the test.

This is the type of action research that would really contribute to the profession…

<span style="background-color: #ffffff; color: #000000; font-family: Calibri,sans-serif; font-size: 12pt;">John